they may be difficult to collect because of a verti- 

 cally stratified distribution or they may remain 

 unrecognized in midwater plankton samples be- 

 cause of their fragility, transparency, and devia- 

 tion from typical larvacean structure. 



Epipelagic larvacean relatives of B. charon 

 filter feed on nanoplankton, especially cells <10 

 Mm (Lohmann 1899; Alldredge 1975). The muscu- 

 lar tail pumps water through the house and per- 

 mits the concentration of suspended particles from 

 larger volumes of water than would be possible 

 using ciliary currents alone. Since food is selected 

 only on the basis of size, detritus may constitute a 

 significant fraction of the food in some locations 

 (Gerber and Marshall 1974). 



In waters below the euphotic zone, particulate 

 organic carbon is scarce, generally present at 

 levels from 10 to 10^ ixg C/1, compared with 

 roughly 10^ to 10^ /xg C/1 in the euphotic zone 

 (Holm-Hansen et al. 1966; Hobson 1967; Menzel 

 1967). Most of the particulate carbon below 200 m 

 contains little or no chlorophyll (Holm-Hansen et 

 al. 1966) and is composed mainly of detritus. How- 

 ever, Fournier ( 1971 ) and others have reported the 

 presence of living, pigmented cells ("olive-green 

 cells," or OGC's) averaging 3.5 fj-m in diameter in 

 virtually all waters sampled deeper than about 50 

 m in the Atlantic and Pacific Oceans. These cells 

 reach their maximum density of about 10^/1 at 300 

 to 500 m and may contribute up to about 1 /Lig C/1, 

 or up to 10% of the total particulate organic carbon 

 in aphotic marine environments. Fournier (1971) 

 suggested that copepods are not likely to be major 

 consumers of OGC's because of their limited 

 abilities to filter such small particles at low con- 

 centrations and that pelagic tunicates, which 

 filter water through mucous sheets, may be better 

 suited to utilize such particles. Fournier (1973) 

 demonstrated that the gut contents of colonies of 

 the pelagic tunicate, Pyrosoma, from below the 

 euphotic zone consisted mainly of OGC's. If B. 

 charon is indeed a resident of midwaters, as 

 suggested above, and if it, like its epipelagic rela- 

 tives, filters particles < 10 to 20 /am in size, then it 

 may be a major consumer of OGC's as well as 

 detritus. The localized occurrence of dense layers 

 of S. charon indicated by the in situ observations 

 of Barham ( 1969) may depend on the presence of 

 peak concentrations of OGC's between 200 and 

 1,000 m, as observed by Fournier ( 1971). Alterna- 

 tively, the filter meshes of the house of B. charon 

 may be larger than those of smaller, epipelagic 

 larvaceans and the food may then consist largely 



of slow zooplankton. Knowledge of house structure 

 and analyses of gut contents of additional speci- 

 mens may clarify the role of B. charon in meso- 

 pelagic food webs. 



Bathochordaeiis charon may contribute large 

 amounts of mucus to the water column in the form 

 of its discarded houses. Oikopleura clioica secretes 

 and discards four to six houses per day (Paf- 

 fenhofer 1973). Such occupied and empty houses 

 are sources of particulate food and surface habitat 

 for microorganisms in planktonic ecosystems 

 (Alldredge 1972, 1976a) and, along with other or- 

 ganic aggregates, may serve as a barrier to the 

 downward flux of particulate matter and sub- 

 stances adherred or adsorbed to them ( Silver et al. 

 1978). Moreover, such "marine snow" provides a 

 trophic link between large consumers and nano- 

 plankton, protozoa, and microcrustaceans, allow- 

 ing the former to tap an otherwise unavailable 

 food source (Hamner et al. 1975). Larvaceans and 

 their houses are known prey for fish and 

 planktonic invertebrates (Alldredge 1976a, b; 

 Bailey et al. 1975; Hobson 1974; Hobson and Chess 

 1976). 



Bathochordaeiis charon produces large mucous 

 structures, although the size and frequency of pro- 

 duction of the houses is not known. The rate of 

 turnover of houses is probably less than in O. 

 clioica because of lower temperatures and lower 

 concentrations of particulates which could clog the 

 house filters. Other Oikopleuridae produce houses 

 which are roughly 5 to 15 times the trunk length 

 (Alldredge 1975). If this ratio holds forB. charon, 

 then a 25 mm individual would produce a house 

 about 10 to 40 cm in diameter, comparable with in 

 situ estimates (Barham 1969). IfB. charon is con- 

 centrated in layers just above the thermocline, as 

 suggested by in situ observations (Barham see 

 footnote 2), then its houses may form a major com- 

 ponent of mesopelagic marine snow. 



Note Added in Proof 



I am grateful to A. Biickmann and H. Kapp for 

 calling my attention to their paper (Unter- 

 suchungen am Zooplankton von der Atlantischen 

 Kuppenfahrt der „Meteor", Marz bis Juli 1967, 

 published 1973 in „ Meteor" Forschungsergeb- 

 nisse, Reihe D, No. 13:11-36) in which they de- 

 scribed and illustrated two additional specimens, 

 referred to as B. stygius. The specimens were 

 taken April 1967 in the North Atlantic (lat. 30°18' 

 N, long. 29^20' W) between 100 m and the surface. 



518 



